R Technology Workshop

R is the most popular free software environment for statistical computing and graphics. ggplot2 is a data visualization package for R that can be used to produce publication-quality graphics. This workshop is designed to introduce you to R and ggplot as well as RStudio, KnitR, Slidify, and Shiny.
R is a central piece of the Big Data Analytics Revolution, for example, see http://opensource.com/business/14/7/interview-david-smith-revolution-analytics for an article entitled “Big data influencer on how R is paving the way”

This is how my RStudio is configured:

sessionInfo()

## R version 3.1.1 (2014-07-10)
## Platform: x86_64-apple-darwin13.3.0 (64-bit)
## 
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## loaded via a namespace (and not attached):
## [1] digest_0.6.4     evaluate_0.5.5   formatR_1.0      htmltools_0.2.4 
## [5] knitr_1.6        rmarkdown_0.2.64 stringr_0.6.2    tools_3.1.1     
## [9] yaml_2.1.13

You also need to install LaTeX if you want to generate PDF files from KnitR.

http://latex-project.org/ftp.html

Getting Started - Clone the RWorkshop GiT Repository:

Phils-MacBook-Pro:Mine pcannata$ pwd
/Users/pcannata/Mine
Phils-MacBook-Pro:Mine pcannata$ git clone https://github.com/pcannata/RWorkshop.git
Cloning into ‘RWorkshop’…
remote: Counting objects: 19, done.
remote: Compressing objects: 100% (12/12), done.
remote: Total 19 (delta 3), reused 19 (delta 3)
Unpacking objects: 100% (19/19), done.
Checking connectivity… done
Phils-MacBook-Pro:Mine pcannata$ ls -a RWorkshop/
. .Rprofile.R 00 Doc 03 ggplot 05 KnitR 02
.. .git 01 Basic R 04 KnitR 01 RWorkshop.Rproj

Getting Started - Create a New RStudio Project for the code in the cloned repository:

Getting Started - Create .Rprofile file to load ggplot and gplot libraries when the project is started:

Create an new file named .Rprofile by copying a file, see below

Copy what’s in blue above into the new file named .Rprofile as below,

Recommended Books:

http://www.amazon.com/Everyone-Advanced-Analytics-Graphics-Addison-Wesley/dp/0321888030/ref=sr_1_1?ie=UTF8&qid=1404611818&sr=8-1&keywords=R+for+everyone

http://www.amazon.com/Reproducible-Research-RStudio-Chapman-Series/dp/1466572841/ref=sr_1_1?ie=UTF8&qid=1404612129&sr=8-1&keywords=reproducible+research+with+r

Basic R Language Constructs

See also http://cran.r-project.org/doc/manuals/r-devel/R-lang.html, http://www.r-tutor.com/r-introduction, and http://www.cookbook-r.com/

source("../01 Basic R/Basic.R", echo = TRUE)

## 
## > "Variables"
## [1] "Variables"
## 
## > v <- 211
## 
## > v
## [1] 211
## 
## > "Global Variables"
## [1] "Global Variables"
## 
## > g <<- 234
## 
## > g
## [1] 234
## 
## > "Vectors"
## [1] "Vectors"
## 
## > v1 <- c(1, 2, 3, 4, 5)
## 
## > v1
## [1] 1 2 3 4 5
## 
## > v2 <- 1:11
## 
## > v2
##  [1]  1  2  3  4  5  6  7  8  9 10 11
## 
## > v3 <- -5:5
## 
## > v3
##  [1] -5 -4 -3 -2 -1  0  1  2  3  4  5
## 
## > "Vector Operations"
## [1] "Vector Operations"
## 
## > v1
## [1] 1 2 3 4 5
## 
## > v1 + 2
## [1] 3 4 5 6 7
## 
## > v2
##  [1]  1  2  3  4  5  6  7  8  9 10 11
## 
## > sqrt(v2)
##  [1] 1.000 1.414 1.732 2.000 2.236 2.449 2.646 2.828 3.000 3.162 3.317
## 
## > v2
##  [1]  1  2  3  4  5  6  7  8  9 10 11
## 
## > v3
##  [1] -5 -4 -3 -2 -1  0  1  2  3  4  5
## 
## > v2 + v3
##  [1] -4 -2  0  2  4  6  8 10 12 14 16
## 
## > length(v3)
## [1] 11
## 
## > mean(4:22)
## [1] 13
## 
## > "Data Types: Numeric, Character, Dates, Logical(TRUE, FALSE)"
## [1] "Data Types: Numeric, Character, Dates, Logical(TRUE, FALSE)"
## 
## > "Missing Data: NA"
## [1] "Missing Data: NA"
## 
## > v <- c(1, 2, NA, 3)
## 
## > v
## [1]  1  2 NA  3
## 
## > "Missing Data: NULL"
## [1] "Missing Data: NULL"
## 
## > v <- c(1, 2, NULL, 3)
## 
## > v
## [1] 1 2 3
## 
## > "Functions"
## [1] "Functions"
## 
## > "Functions will be introduced in the section pn ggplot below, however, let's have a look at the apropos() function:"
## [1] "Functions will be introduced in the section pn ggplot below, however, let's have a look at the apropos() function:"
## 
## > apropos("mean")
##  [1] ".colMeans"     ".rowMeans"     "colMeans"      "kmeans"       
##  [5] "mean"          "mean.Date"     "mean.default"  "mean.difftime"
##  [9] "mean.POSIXct"  "mean.POSIXlt"  "rowMeans"      "weighted.mean"
## 
## > "Data Structures: Dataframes, Lists, Matricies, and Arrays. Only Dataframes will be addressed in this workshop."
## [1] "Data Structures: Dataframes, Lists, Matricies, and Arrays. Only Dataframes will be addressed in this workshop."

R Dataframes

A data frame is used for storing data tables. It is a list of vectors of equal length. For example, the following variable df is a data frame containing three vectors n, s, b.

n = c(2, 3, 5) 
s = c("aa", "bb", "cc") 
b = c(TRUE, FALSE, TRUE) 
df = data.frame(n, s, b)       # df is a data frame
head(df)

##   n  s     b
## 1 2 aa  TRUE
## 2 3 bb FALSE
## 3 5 cc  TRUE

Dataframes can be loaded from databases, CSVs, Excel, etc.. Loading dataframes from an Oracle database will be discussed later in this Workshop.

See also http://www.r-tutor.com/r-introduction/data-frame

Many R packages come with demo dataframes. The ggplot package comes with a demo dataframe called diamonds which we will use for this workshop.

source("../02 R Dataframes/Dataframes.R", echo = TRUE)

## 
## > library("ggplot2")
## 
## > "Displaying the top few rows of a dataframe:"
## [1] "Displaying the top few rows of a dataframe:"
## 
## > head(diamonds)
##   carat       cut color clarity depth table price    x    y    z
## 1  0.23     Ideal     E     SI2  61.5    55   326 3.95 3.98 2.43
## 2  0.21   Premium     E     SI1  59.8    61   326 3.89 3.84 2.31
## 3  0.23      Good     E     VS1  56.9    65   327 4.05 4.07 2.31
## 4  0.29   Premium     I     VS2  62.4    58   334 4.20 4.23 2.63
## 5  0.31      Good     J     SI2  63.3    58   335 4.34 4.35 2.75
## 6  0.24 Very Good     J    VVS2  62.8    57   336 3.94 3.96 2.48
## 
## > "Selecting a subset of columns from a dataframe:"
## [1] "Selecting a subset of columns from a dataframe:"
## 
## > head(subset(diamonds, select = c(carat, cut)))
##   carat       cut
## 1  0.23     Ideal
## 2  0.21   Premium
## 3  0.23      Good
## 4  0.29   Premium
## 5  0.31      Good
## 6  0.24 Very Good
## 
## > "Selecting a subset of rows from a dataframe:"
## [1] "Selecting a subset of rows from a dataframe:"
## 
## > head(subset(diamonds, cut == "Ideal" & price > 5000))
##       carat   cut color clarity depth table price    x    y    z
## 11417  1.16 Ideal     E     SI2  62.7  56.0  5001 6.69 6.73 4.21
## 11418  1.16 Ideal     E     SI2  59.9  57.0  5001 6.80 6.82 4.08
## 11422  1.07 Ideal     I     SI1  61.7  56.1  5002 6.57 6.59 4.06
## 11423  1.10 Ideal     H     SI2  62.0  56.5  5002 6.58 6.63 4.09
## 11424  1.20 Ideal     J     SI1  62.1  55.0  5002 6.81 6.84 4.24
## 11431  1.14 Ideal     H     SI1  61.6  57.0  5003 6.70 6.75 4.14
## 
## > "Find average price group by color (plyr package is needed)"
## [1] "Find average price group by color (plyr package is needed)"
## 
## > library("plyr")
## 
## > ddply(subset(diamonds, cut == "Ideal" & price > 5000), 
## +     ~color, summarise, o = mean(price, na.rm = TRUE))
##   color    o
## 1     D 9057
## 2     E 9065
## 3     F 9704
## 4     G 9392
## 5     H 8923
## 6     I 9663
## 7     J 9407

For more on subsetting dataframes see http://www.ats.ucla.edu/stat/r/faq/subset_R.htm

Connecting to Oracle with RJDBC

RJDBC is an R package for makeing database connections in R.

See also http://www.rforge.net/RJDBC/, and http://bommaritollc.com/2012/11/connecting-r-to-an-oracle-database-with-rjdbc/

source("../02A RJDBC/ConnectToOracle.R", echo = TRUE)

## 
## > "\nPut the following into .bash_profile \nexport JAVA_HOME=`/usr/libexec/java_home` \n. ./.bash_profile \n\nDownload ojdbc6.jar into ~/Downloads \ns ..." ... [TRUNCATED] 
## [1] "\nPut the following into .bash_profile \nexport JAVA_HOME=`/usr/libexec/java_home` \n. ./.bash_profile \n\nDownload ojdbc6.jar into ~/Downloads \nsudo mv ~/Downloads/ojdbc6.jar $JAVA_HOME \n"
## 
## > Sys.setenv(JAVA_HOME = "/Library/Java/JavaVirtualMachines/jdk1.7.0_60.jdk/Contents/Home")
## 
## > options(java.parameters = "-Xmx2g")
## 
## > library(rJava)
## 
## > .jinit()
## 
## > print(.jcall("java/lang/System", "S", "getProperty", 
## +     "java.version"))
## [1] "1.7.0_60-ea"
## 
## > library(RJDBC)

## Loading required package: DBI

## 
## > jdbcDriver <- JDBC(driverClass = "oracle.jdbc.OracleDriver", 
## +     classPath = "~/ojdbc6.jar")
## 
## > e1 <- 7369
## 
## > e2 <- "SMITH"
## 
## > e3 <- "CLERK"
## 
## > e4 <- 7902
## 
## > e5 <- "17-DEC-1980"
## 
## > e6 <- 800
## 
## > e7 <- 20
## 
## > emps <- data.frame(e1, e2, e3, e4, e5, e6, e7)
## 
## > col_headings <- c("EMPNO", "ENAME", "JOB", "MGR", 
## +     "HIREDATE", "SAL", "DEPTNO")
## 
## > names(emps) <- col_headings
## 
## > possibleError <- tryCatch(jdbcConnection <- dbConnect(jdbcDriver, 
## +     "jdbc:oracle:thin:@zenji.microlab.cs.utexas.edu:1521:orcl", 
## +     "C##cs34 ..." ... [TRUNCATED] 
## 
## > if (!inherits(possibleError, "error")) {
## +     emps <- dbGetQuery(jdbcConnection, "select * from emp")
## +     dbDisconnect(jdbcConnection)
## + }
## [1] TRUE
## 
## > ggplot(data = emps) + geom_histogram(aes(x = SAL))

## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.

ggplot2

ggplot is an R package for data exploration and visualization. It produces production quality graphics and allows you to slice and dice your data in many different ways. ggplot uses a general scheme for data visualization which breaks graphs up into semantic components such as scales and layers. In contrast to other graphics packages, ggplot2 allows the user to add, remove or alter components in a plot at a high level of abstraction.

See also http://ggplot2.org/, http://cran.r-project.org/web/packages/ggplot2/ggplot2.pdf, and https://groups.google.com/forum/#!forum/ggplot2

source("../03 ggplot/Plots.R", echo = TRUE)

## 
## > options(java.parameters = "-Xmx2g")
## 
## > jdbcDriver <- JDBC(driverClass = "oracle.jdbc.OracleDriver", 
## +     classPath = "~/ojdbc7.jar")
## 
## > con <- dbConnect(jdbcDriver, "jdbc:oracle:thin:@128.83.138.158:1521:orcl", 
## +     "c##cs347_zi322", "orcl_zi322")
## 
## > HCAHPSMeasure = dbGetQuery(con, "Select * From mc_HCAHPSMeasure")
## 
## > InpatientServices <- dbGetQuery(con, "Select * from mc_InpatientServices")
## 
## > Providers = dbGetQuery(con, "Select * from mc_Providers")
## 
## > OutpatientServices <- dbGetQuery(con, "Select * from mc_OutpatientServices")
## 
## > OutpatientVisits <- dbGetQuery(con, "select * from mc_OutpatientVisits_2 WHERE ID BETWEEN 0 and 30000")
## 
## > OutpatientVisits = rbind(OutpatientVisits, dbGetQuery(con, 
## +     "select * from mc_OutpatientVisits_2 WHERE ID BETWEEN 30001 and 60000"))
## 
## > OutpatientVisits = rbind(OutpatientVisits, dbGetQuery(con, 
## +     "select * from mc_OutpatientVisits_2 WHERE ID BETWEEN 60001 and 90000"))
## 
## > OutpatientVisits = rbind(OutpatientVisits, dbGetQuery(con, 
## +     "select * from mc_OutpatientVisits_2 WHERE ID BETWEEN 90001 and 100000"))
## 
## > InpatientVisits <- dbGetQuery(con, "select * from mc_InpatientVisits WHERE ID BETWEEN 0 and 30000")
## 
## > InpatientVisits = rbind(InpatientVisits, dbGetQuery(con, 
## +     "select * from mc_InpatientVisits WHERE ID BETWEEN 30001 and 60000"))
## 
## > InpatientVisits = rbind(InpatientVisits, dbGetQuery(con, 
## +     "select * from mc_InpatientVisits WHERE ID BETWEEN 60001 and 90000"))
## 
## > InpatientVisits = rbind(InpatientVisits, dbGetQuery(con, 
## +     "select * from mc_InpatientVisits WHERE ID BETWEEN 90001 and 100000"))
## 
## > InpatientVisits = rbind(InpatientVisits, dbGetQuery(con, 
## +     "select * from mc_InpatientVisits WHERE ID BETWEEN 100001 and 130000"))
## 
## > InpatientVisits = rbind(InpatientVisits, dbGetQuery(con, 
## +     "select * from mc_InpatientVisits WHERE ID BETWEEN 130001 and 160000"))
## 
## > head(InpatientVisits)
##    ID DRGID PROVIDERID NUMDISCHARGEDPATIENTS COVEREDCHARGES TOTALPAYMENTS
## 1 666    39     340073                    14          24347   5770.785714
## 2 667    39     340091                    67          13557   6375.268657
## 3 668    39     340109                    17          15891   7125.058824
## 4 669    39     340113                    60          43530   8805.083333
## 5 670    39     340114                    26          21445   6177.230769
## 6 671    39     340115                    55          16703   6198.163636
##   MEDICAREPAYMENT YEAR
## 1            4785 2012
## 2            5289 2012
## 3            6112 2012
## 4            6573 2012
## 5            4427 2012
## 6            5299 2012
## 
## > outpatientCostByCity = dbGetQuery(con, "SELECT mc_Providers.City as City, AVG(mc_OutPatientVisits.AverageSubmittedCharges) as AvgBilledCost \nFROM m ..." ... [TRUNCATED] 
## 
## > outpatientCostByState = dbGetQuery(con, "SELECT mc_Providers.State as State, AVG(mc_OutPatientVisits.AverageSubmittedCharges) as AvgBilledCost \nFRO ..." ... [TRUNCATED] 
## 
## > outpatientCostByHospital = dbGetQuery(con, "\nSELECT mc_Providers.Name as Hospital, AVG(mc_OutPatientVisits.AverageSubmittedCharges) as AvgBilledCos ..." ... [TRUNCATED] 
## 
## > outpatientCostByCity = dbGetQuery(con, "SELECT mc_Providers.City as City, AVG(mc_OutPatientVisits.AverageSubmittedCharges) as AvgBilledCost \nFROM m ..." ... [TRUNCATED] 
## 
## > outpatientCostByState = dbGetQuery(con, "SELECT mc_Providers.State as State, AVG(mc_OutPatientVisits.AverageSubmittedCharges) as AvgBilledCost \nFRO ..." ... [TRUNCATED] 
## 
## > outpatientCostByHospital = dbGetQuery(con, "SELECT mc_Providers.Name as Hospital, AVG(mc_OutPatientVisits.AverageSubmittedCharges) as AvgBilledCost  ..." ... [TRUNCATED] 
## 
## > InpatientCostByCity = dbGetQuery(con, "SELECT mc_Providers.City as City, AVG(mc_InPatientVisits.CoveredCharges) as AvgBilledCost \nFROM mc_InPatient ..." ... [TRUNCATED] 
## 
## > InpatientCostByState = dbGetQuery(con, "SELECT mc_Providers.State as State, AVG(mc_InpatientVisits.CoveredCharges) as AvgBilledCost \nFROM mc_Inpati ..." ... [TRUNCATED] 
## 
## > InpatientCostByHospital = dbGetQuery(con, "\nSELECT mc_Providers.Name as Hospital, AVG(mc_InPatientVisits.CoveredCharges) as AvgBilledCost \nFROM mc ..." ... [TRUNCATED] 
## 
## > p1 <- ggplot(InpatientCostByState, aes(x = STATE, 
## +     y = AVGBILLEDCOST)) + geom_point() + coord_flip()
## 
## > p2 <- ggplot(outpatientCostByState, aes(x = STATE, 
## +     y = AVGBILLEDCOST)) + geom_point() + coord_flip()

The Chapter 7 of “R for Everyone” has many more examples of ggplots.

ggplot2 and functions

source("../03 ggplot/plotFunction.R", echo = TRUE)

## 
## > FigureNum <<- 0
## 
## > ggplot_func <- function(df, Title = "Diamonds", Legend = "color", 
## +     PointColor = c("red", "blue", "green", "yellow", "grey", 
## +         "black" .... [TRUNCATED] 
## 
## > p1 <- ggplot_func(diamonds)

## Scale for 'x' is already present. Adding another scale for 'x', which will replace the existing scale.

## 
## > p1

## 
## > p2 <- ggplot_func(diamonds, YMin = 5000, YMax = 15000)

## Scale for 'x' is already present. Adding another scale for 'x', which will replace the existing scale.

## 
## > p2

## Warning: Removed 40868 rows containing missing values (geom_point).

## 
## > p3 <- ggplot_func(subset(diamonds, cut == "Premium"), 
## +     Legend = "cut")

## Scale for 'x' is already present. Adding another scale for 'x', which will replace the existing scale.

## 
## > p3

## 
## > p4 <- ggplot_func(diamonds, Legend = "clarity", PointColor = c("red", 
## +     "blue", "green", "yellow", "grey", "black", "purple", "orange"))

## Scale for 'x' is already present. Adding another scale for 'x', which will replace the existing scale.

## 
## > p4

## 
## > library("grid", lib.loc = "/Library/Frameworks/R.framework/Versions/3.0/Resources/library")
## 
## > png("4diamonds.png", width = 25, height = 20, units = "in", 
## +     res = 72)
## 
## > grid.newpage()

## 
## > pushViewport(viewport(layout = grid.layout(2, 2)))
## 
## > print(p1, vp = viewport(layout.pos.row = 1, layout.pos.col = 1))
## 
## > print(p2, vp = viewport(layout.pos.row = 1, layout.pos.col = 2))

## Warning: Removed 40868 rows containing missing values (geom_point).

## 
## > print(p3, vp = viewport(layout.pos.row = 2, layout.pos.col = 1))
## 
## > print(p4, vp = viewport(layout.pos.row = 2, layout.pos.col = 2))
## 
## > dev.off()
## pdf 
##   2

You should now be able to open RWorkshop/00 Doc/4diamonds.png. It should look like the following plot.

KnitR

KnitR is an R package designed to generate dynamic reports using a mix of the R, LaTex, and the Rmarkdown (see http://rmarkdown.rstudio.com/?version=0.98.945&mode=desktop) languages.

See also http://yihui.name/knitr/ and http://kbroman.github.io/knitr_knutshell/

Simple examples can be found in “04 KnitR/doc1.Rmd” and “04 KnitR/doc2.Rmd”. These can generate html, pdf, and word documents. The output from Kniting doc2.Rmd is,

A comprehensive KnitR example (which generated this document) can be found in “00 Doc/RWorkshop.Rmd”.

slidify

You can use Slidify to generate HTML slide decks using only the Rmarkdown language.

See also http://slidify.org and http://slidify.org/start.html

Follow the instructions in “05 Slidify/slidify setup.R” to install and run slidify. You should be able to produce a slide deck with a first slide that looks something like the following.

Cool trick - Any github repo with a branch called gh-pages will get served as a website. If the content of that repo is the stuff of websites (html,css), then you get free web hosting. So, create a branch called gh-pages and push to it.

shiny

The shiny R package allows you to build interactive web-based applications using only R with no knowledge of html, css, or javascript needed. You just need to write two scripts (see the example files in the 06Shiny directory):

• ui.R : Defines the layout and the interactive elements that the user can access.
  
• server.R : Defines what computations are done in response to user interactions.

See also http://shiny.rstudio.com and http://shiny.rstudio.com/tutorial

To run the shiny app that’s in the 06Shiny directory run the following in the main RWorkshop directory (make sure the working directory is set to this directory):
library(shiny)
runApp(“06Shiny”) # Make sure there are no spaces in the string argument to runAPP

This should pop the application up in a browser, you can also access it in a browser at http://127.0.0.1:6837. It should look like the following.

shinyapps

The example above ran the shiny app on your local machine, but to share with others, you have to send around the R files and the user needs to have R and know a little bit about it.

Instead, you can remotely host shiny apps and then just send people links. Get a free account at shinyapps.io/signup.html and give it a try.

library(“devtools”, lib.loc=“/Library/Frameworks/R.framework/Versions/3.0/Resources/library”)
install_github( repo = “shinyapps”, username=“rstudio” )
shinyapps::setAccountInfo(name=‘pcannata’, token=‘3ECF447A741004F6A8B7208C9ED778E1’, secret=‘. . .’)

library(shinyapps)
getwd()

## [1] "/Users/alyshialedlie/RWorkshop/00 Doc"

# Uncomment the following line to deploy the app.
#deployApp("../06Shiny")

Now you can try the app at https://pcannata.shinyapps.io/06Shiny/

See also https://www.shinyapps.io/ and http://shiny.rstudio.com/articles/shinyapps.html